INDEX

Explanations

phrases indicating tension or conflict between different entities

oai_token-act-pair · gpt-3.5-turbo

phrases that discuss conflicts or interactions between different groups or entities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bye

-0.96

nit

-0.81

OGR

-0.79

factor

-0.75

nell

-0.72

ãĤ§

-0.70

ificantly

-0.70

jong

-0.70

\\\\\\\\

-0.70

oultry

-0.69

POSITIVE LOGITS

 sexes

1.24

 genders

1.12

 factions

0.91

two

0.88

 combatants

0.86

 halves

0.85

 ourselves

0.85

 spouses

0.84

 them

0.83

 sides

0.83

Activations Density 0.074%

phrases indicating tension or conflict between different entities

phrases that discuss conflicts or interactions between different groups or entities

No Comments

No Known Activations